Taxonomic Survey of Culture-Enriched Metagenomes from Gram-Negative Bacteria in Pigs, Piggery Environments, and Humans Associated with Pig Production Systems in Okada, Edo State, Nigeria

1. Introduction

Pig production constitutes an important component of livestock agriculture and contributes to food supply, income generation, employment and household livelihoods in many parts of the world. The increasing demand for animal protein has encouraged the expansion and intensification of pig production, resulting in the concentration of relatively large numbers of animals within restricted spaces. Under such conditions, pigs, farm workers, feed, water, manure, housing facilities and surrounding environmental compartments are brought into frequent contact [1]. These interactions create a complex microbial ecosystem in which microorganisms can circulate between animals, humans and the environment.

The microbial ecology of pig production systems has received increasing attention because livestock production can provide conditions that support the persistence and dissemination of microorganisms of veterinary, environmental and public-health importance. In particular, bacteria associated with pigs and their production environments may include commensals, opportunistic organisms and recognized pathogens. Gram-negative bacteria are especially important because several members of this group, including species belonging to Escherichia, Klebsiella, Enterobacter, Salmonella, Proteus, Pseudomonas and Acinetobacter, have been associated with animal disease, environmental contamination, and oodborne infections [2].

The pig production environment is particularly relevant in this regard because pigs continuously release microorganisms through faeces and other biological materials. Faecal contamination can subsequently reach floors, pens, drainage channels, wastewater, soil, equipment, footwear and other farm surfaces. Wastewater and manure may further facilitate the movement of microorganisms beyond the immediate farm environment. Scicchitano et al.[3], using shotgun metagenomic sequencing of 294 samples from a swine production system, demonstrated extensive microbial connectivity between pigs and their internal and external farm environments. Their study identified 530 species-level genome bins, of which 309 were shared between animal gut microbiomes and farm-associated environments, illustrating the extent to which microorganisms can occur across different compartments of a pig production system.

Environmental compartments associated with pig production can therefore serve as important reservoirs of microorganisms. Manure, wastewater, soil, dust and contaminated surfaces may support the survival and redistribution of bacteria after their release from animals. This observation is particularly important for Gram-negative bacteria because several members of this group are environmentally persistent and may harbour antimicrobial resistance genes. Previous metagenomic investigations of pig farms have shown that microbial community composition and resistance determinants can extend beyond animal housing into surrounding soil and water. For example, a study of pig-farm wastewater and surrounding environments reported substantial concentrations of resistance genes and identified Proteobacteria, Actinobacteria and Acidobacteria among important bacterial groups associated with the detected resistome [4].

The movement of microorganisms within a pig production system may also involve humans. Farm workers and animal handlers have repeated contact with pigs, faecal material, feed, water, equipment and contaminated surfaces. Such occupational contact creates opportunities for microorganisms to be exchanged between humans and animals. A recent metagenomic study [5] involving swine farms found that farm workers had greater abundances of certain high-risk antimicrobial resistance genes than non-exposed human controls and identified occupational contact as an important interface in the distribution of resistance-associated microorganisms. The human–animal–environment interface is therefore an important consideration when investigating the microbial ecology of pig production systems [5]. Microorganisms detected in humans may originate from occupational exposure, community sources or other routes, and detection of the same organism in different compartments does not by itself demonstrate transmission. Nevertheless, identifying shared bacterial taxa among pigs, farm environments and humans provides useful information for understanding microbial connectivity and identifying organisms that warrant further investigation.

Conventional microbiological techniques have traditionally been used to investigate bacteria associated with animals and their environments [6]. Culture and biochemical characterization remain valuable approaches because they allow viable organisms to be recovered and characterized. However, culture-dependent techniques do not necessarily capture the full diversity of microorganisms present in a sample. Some bacteria may occur at low abundance, require specific growth conditions or be outcompeted by more readily cultivable organisms. Consequently, dependence on conventional culture alone can result in an incomplete representation of the microbial community. Metagenomic sequencing has expanded the capacity to investigate microbial communities by allowing genetic material from multiple organisms to be analyzed simultaneously. Shotgun metagenomic sequencing is particularly useful for microbial community analysis because it does not depend on amplification of a single taxonomic marker. Instead, genomic DNA from the microbial community is sequenced, allowing sequences originating from different organisms to be analyzed simultaneously [3]. This approach can provide information at multiple taxonomic levels and, when sequence quality and reference databases are adequate, can support species-level classification and genome reconstruction. Culture-enriched metagenomics provides a complementary strategy in which microbial enrichment or cultivation is incorporated before sequencing. The approach can increase the representation of organisms of interest and facilitate genomic characterization of organisms that may occur at low abundance in the original sample [7]. In the context of pig production systems, this approach is particularly useful when the objective is to investigate specific bacterial groups, such as Gram-negative bacteria, while retaining the broader taxonomic information obtainable through sequencing.

Taxonomic characterization is one of the fundamental applications of metagenomic sequencing. Taxonomic analysis involves assigning DNA sequences, assembled contigs or reconstructed genomes to appropriate taxonomic categories [3]. Depending on the quality of the sequence data and the availability of reference genomes, microorganisms can be assigned to levels ranging from phylum and class to order, family, genus and species. Such information allows investigators to determine which bacterial groups occur within particular environments and to compare their distribution among different sample types. The taxonomic composition of Gram-negative bacteria in pig production systems is of particular interest because these organisms occupy diverse ecological niches. Members of the Enterobacterales, for example, are frequently associated with gastrointestinal environments but can also occur in soil, water, manure and contaminated farm surfaces [8]. Other Gram-negative organisms, including Pseudomonas and Acinetobacter, can persist in environmental habitats and may possess characteristics that enable them to survive under adverse conditions. The simultaneous occurrence of these organisms across animal and environmental compartments can provide useful information about the microbial ecology of livestock production.

Recent metagenomic research has strengthened the value of taxonomic analysis across pig production systems. Scicchitano et al. (2024) demonstrated that shotgun metagenomics could identify hundreds of species-level genome bins across pigs, farm environments, surrounding soil, wastewater and slaughterhouse-associated sites. Their observation that hundreds of genome bins were shared between animal and environmental compartments highlights the importance of investigating microbial communities across the entire production system rather than restricting surveillance to the animal host alone. Similarly, recent research has extended the investigation to human-associated compartments. A 2025 metagenomic study of a swine-farm ecosystem examined paired samples from pig faeces, farm soil, farm workers and non-exposed human controls and demonstrated differences in high-risk antimicrobial resistance genes between exposed and non-exposed human groups [5]. Although resistance genes were the principal focus of that study, the findings reinforce the importance of considering humans as an integral component of the microbial ecosystem surrounding pig production.

Despite the increasing application of metagenomics in livestock research, significant geographical gaps remain. Much of the published evidence concerning pig-associated microbiomes has been generated in Europe, Asia and other regions with different production systems, environmental conditions and antimicrobial-use practices. The findings from these settings cannot necessarily be assumed to represent pig production systems in Nigeria. Local investigations are therefore necessary to establish the bacterial taxa occurring within Nigerian production environments. This need is particularly relevant in Nigeria, where pig production occurs under diverse management systems ranging from relatively intensive commercial farms to smaller-scale operations. Differences in housing, animal density, sanitation, manure management, water sources, feed, and veterinary support may influence the bacterial communities associated with individual farms. The environmental dimension is equally important. Pig manure and wastewater may introduce microorganisms into surrounding soil and water systems, particularly where waste treatment and disposal are inadequate.

Okada, Edo State, provides an appropriate setting for examining these relationships. The area contains agricultural, residential, commercial and institutional activities, and pig production forms part of the local livestock economy. Pigs, farm workers and environmental materials within production facilities may interact repeatedly, creating opportunities for microorganisms to move among different ecological compartments. However, there remains limited sequence-based information describing the Gram-negative bacterial communities associated with pigs, piggery environments and humans within pig production systems in the area. A taxonomic survey of culture-enriched metagenomes can therefore provide valuable baseline information. By identifying and comparing bacterial taxa across pigs, piggery environments and humans, it becomes possible to determine whether particular taxa are broadly distributed or associated more strongly with specific compartments. Shared taxa may indicate ecological overlap, whereas compartment-specific taxa may reflect differences in environmental conditions, host-associated microbiology or patterns of exposure. It is important, however, to distinguish taxonomic detection from biological interpretation. Detection of a bacterial sequence does not necessarily demonstrate that the organism is viable, pathogenic or responsible for disease. Similarly, the occurrence of the same taxon in two compartments does not establish direct transmission between them. Instead, taxonomic data provide an evidence base from which ecological relationships and potential transmission pathways can be investigated using additional epidemiological, genomic and experimental approaches. The One Health framework provides an appropriate conceptual basis for such investigations because it recognizes the interconnectedness of human, animal and environmental health. Within this framework, microbial surveillance considers organisms and their associated genetic characteristics across multiple ecological compartments rather than treating human, animal and environmental microbiology as completely independent fields. Recent metagenomic studies of pig production systems have increasingly adopted this integrated perspective by analyzing pigs, humans, farm environments and, in some cases, downstream stages of the pork production chain within the same investigative framework [3].

The taxonomic investigation of Gram-negative bacteria is consequently an important first step in understanding the microbial structure of pig production systems. Establishing which organisms are present provides a foundation for examining their ecological distribution and for subsequently investigating antimicrobial resistance, virulence and mobile genetic elements. In a Nigerian context, such information is particularly valuable because local data remain relatively limited compared with the expanding international literature. Against this background, the present study focuses on the taxonomic survey of culture-enriched metagenomes from Gram-negative bacteria in pigs, piggery environments and humans associated with pig production systems in Okada, Edo State, Nigeria. The study seeks to characterize the Gram-negative bacterial taxa represented in culture-enriched metagenomic samples and compare their distribution among pigs, environmental compartments and humans. The resulting information is expected to provide a baseline description of the bacterial diversity associated with the pig production system and contribute to broader One Health investigations of microbial ecology, antimicrobial resistance and potential movement of bacteria across animal–human–environment interfaces.

2. Materials and Methods

2.1. Study Area

The study was conducted in Okada, Ovia South-West Local Government Area of Edo State, Nigeria. Okada is a rural and semi-rural community in Edo State where agricultural activities, including livestock production, form an important component of local livelihoods. Pig production systems in the area comprise animal housing facilities, feeding areas, manure accumulation sites, drainage channels and other environmental compartments that can facilitate interaction between livestock, farm workers and the surrounding environment.

2.2. Study Design

A cross-sectional descriptive study was conducted to characterize Gram-negative bacteria associated with the pig production system at Okada, Edo State, Nigeria. Samples were collected from pigs, pig-associated farm environments and human-associated components of the production system. The sampling framework was designed to capture Gram-negative bacterial populations occurring at different points of contact within the pig production system. The principal sampling units included: pig skin and faecal samples; pig-associated environmental samples such as wastewater, soil, and feeds collected from piggery environments; as well as urine and nasal swab samples of consenting farm workers and other humans in close association with the piggery environment. The bacterial isolates were recovered using standard microbiological procedures and identified using conventional and molecular techniques. Representative Gram-negative isolates from each of the three compartments were pooled to create culture-enriched metagenomic samples that were sequenced using whole-genome shotgun sequencing technique, and genomic data analyzed to determine bacterial identity, thereby providing a One Health assessment of bacterial characterization within the production system.

2.3. Ethics Approval and Consent of Pig Farm Owner

Permission to conduct this study was given by the Ethical Research Committee of the Igbinedion University, Okada, with reference number IUO/Ethics/061/25. The purpose of the study was explained to the pig farm owners and informed oral consent was obtained from them.

2.4. Collection and Transportation of Samples

Sterile containers and sterile swabs were used according to sample type. The samples were collected aseptically to minimize contamination from external sources. Faecal samples were collected from rectal sampling procedures. Environmental samples were collected from relevant sites such as piggery-associated soils, feeds, and drainage or waste-disposal points. Samples were immediately labelled and placed in a cooled transport container. Samples were transported to the microbiology laboratory and processed within four hours of sample collection to preserve bacterial viability and reduce changes in the microbial population.

2.5. Culture Enrichment and Isolation of Gram-Negative Bacteria

Culture enrichment involved inoculating samples into buffered peptone water enrichment broths and incubating them overnight at 35 ^oC suitable for the recovery of Gram-negative bacteria [9,10]. Following enrichment, aliquots were streaked onto Chromocult Tryptone Bile-glucuronide agar medium (Techno Pharmchem Ltd, India) for the primary isolation and differentiation of Gram-negative bacteria. After inoculation, the culture plates were incubated at 35 ^oC for 18 to 24 hours. Following incubation, colonies displaying different morphological characteristics were selected for purification. Colony morphology, pigmentation, lactose fermentation, mucoid appearance, and other observable characteristics were recorded. Distinct colonies were repeatedly subcultured until pure cultures were obtained. Pure isolates were maintained andstored in Nutrient Agar slants (HiMedia, India) at 4 ^oCfor subsequent phenotypic and molecular analyses.

2.6. Identification of Bacterial Isolates

Preliminary identification was performed using standard microbiological procedures [11,12]. Isolates were examined by Gram staining and evaluated using appropriate biochemical tests. The biochemical identification scheme included relevant tests such as: lactose fermentation, indole, citrate utilization, methyl red, and Voges–Proskauer. Results were interpreted using standard bacteriological identification criteria. Because phenotypic identification can occasionally fail to distinguish closely related members of the Enterobacterales and other Gram-negative groups, genomic identification was subsequently used to provide higher-resolution taxonomic assignment.

2.7. Genomic DNA Extraction

Metagenomic DNA of the freshly cultured bacterial broth was extracted using a validated commercial bacterial genomic DNA extraction kit following the manufacturer's instructions (Zymo Research Corporation, USA).

2.8. Assessment of DNA Quantity and Quality

Extracted metagenome was evaluated for both concentration and purity before sequencing. DNA concentration was determined using a fluorometric method, while purity was assessed spectrophotometrically using absorbance ratios. DNA integrity was also evaluated using agarose gel electrophoresis. Only DNA preparations meeting the sequencing provider's or library-preparation manufacturer's quality requirements were advanced to sequencing.

2.9. Whole-Genome Shotgun Library Preparation

Metagenomic DNA was quantified fluorometrically with a Qubit® 2.0 Fluorometer (Thermo Fisher Scientific, Waltham, MA, USA) using a target specific Qubit assay (dsDNA BR Assay Kit, Thermo Fisher Scientific).The Illumina Nextera XT DNA library preparation kit (Illumina Inc, San Diego, CA, USA) was used for preparation of fragment libraries of the bacterial genomes, and index codes were added to attribute sequences to each sample, as previously described [13]. A total of 1ng of genomic DNA per sample was used for DNA fragment library preparation. Briefly, the DNA sample was fragmented by sonication to a size of 350bp, then DNA fragments were end-polished, A-tailed, and ligated with the full-length adaptor for Illumina sequencing, with further PCR amplification. The PCR products were ultimately purified (AMPure XP system) and libraries were analyzed for size distribution by Agilent2100 Bioanalyzer and quantified using real-time PCR. The clustering of the index-coded samples was performed on a cBot Cluster Generation System according to the manufacturer’s instructions.

2.10. Whole-Genome Shotgun Sequencing

The prepared libraries were subjected to whole-genome shotgun sequencing using a high-throughput sequencing platform. Sequencing generated paired-end short reads from the Illumina-type platform used [13,14]. Briefly, the library preparations were sequenced on an Illumina HiSeq platform and paired-end reads (2×150 bp) were generated. Raw sequence data were generated and FASTQ files were obtained. The quality of obtained data was assessed using FASTQC (v.0.11.5) and low-quality reads were filtered out and adapter regions were removed using Trimmomatic (v.0.36) [15]. The raw reads obtained from sequencing were assembled denovo using MEGAHIT version 1.2.9 [16], and the quality of assembled genome was determined through Quality Assessment Tool (QUAST) v5.0.2 [17]. The assembled contigs were deposited in the National Center for Biotechnological Information (NCBI) repository.

2.11. Sequencing Quality Control

Raw sequencing reads were subjected to quality assessment before downstream analysis. Quality-control procedures included assessment of: per-base sequence quality ,per-sequence quality scores, GC content, sequence length distribution, adapter contamination, duplicated reads, and other indicators of sequencing quality. FastQC was used for initial read-level quality. FastQC produces detailed quality reports for high-throughput sequencing datasets, allowing problematic libraries to be identified before downstream analysis. Low-quality bases and sequencing adapters were removed using an appropriate read-trimming program [Trimmomatic (v.0.36)]. Reads that remained below the predetermined quality or length threshold after trimming were discarded.

2.12. De novo Genome Assembly

High-quality sequencing reads were assembled de novo to reconstruct draft genomes. A short-read assembler (MEGAHIT version 1.2.9) appropriate for metagenomic WGS data was used. Assembly quality was assessed using parameters including: total assembly length, number of contigs, N50, L50, largest contig, and GC content. Assemblies with substantial contamination, excessive fragmentation or otherwise inadequate quality were excluded from downstream comparative analyses or subjected to additional processing.

2.13. Taxonomic Classification of Sequencing Reads

Taxonomic classification was performed to evaluate the composition and identity of the bacterial sequence data. Since the culture-enriched samples contained mixed sequence reads, a k-mer-based taxonomic classifier, Kraken 2, was used to classify reads against a curated reference database. Bracken was then used to estimate taxonomic abundance from Kraken-classified reads. Because the principal objective was characterization of cultured bacterial isolates, taxonomic classification was subsequently reconciled with the original culture-based identification.

2.14. Statistical Analysis

Phenotypic data were entered into a spreadsheet and subsequently analyzed using Number Cruncher Statistical Software (NCSS) ver. 12 data analysis software. Descriptive statistics were used to summarize frequency of bacterial species while Pearson chi-square test of independence, goodness-of-fit analysis, Cochran's Q and Jaccard tests were used to evaluate the bacterial distribution. Statistical significance was set at p < 0.05.

3. Results

3.1. Distribution of Samples Examined From the Pig Production System

A total of 356 samples were examined from the pig production system at Okada, Edo State, representing the three principal components of the One Health framework: animal, environmental and human-associated sources (Table 1). Pig-associated samples constituted the largest proportion of the samples collected, with 132 samples (37.08%) obtained from faecal, rectal and skin sources. Environmental samples accounted for 124 samples (34.83%), comprising wastewater, soil and feed, while 100 human-associated samples (28.09%) were obtained from farm workers and other individuals who had close contact with the piggery environment. The relatively close distribution of samples among the three compartments indicates that the investigation did not focus exclusively on pigs but incorporated the wider ecological setting in which pig production occurs. Animal samples represented the principal biological reservoir, whereas environmental samples captured potential pathways through which microorganisms and resistance determinants could persist or circulate within the production system. Human-associated samples provided an additional component for assessing the interface between occupational exposure and the piggery environment.

3.2. Phenotypic Identification and Frequency of Occurrence of Gram-Negative Bacteria

Following cultural examination, Gram staining and biochemical characterization (Table 2), Fourty-eight (48) Gram-negative bacterial isolates were recovered from the pig production system. The isolates originated from the three One Health compartments: pig-associated samples, environmental samples and human-associated samples. Pig-associated samples yielded the highest number of isolates (22; 45.83%), followed by environmental samples (14; 29.17%) and human-associated samples (12; 25.00%).The isolates were phenotypically characterized based on colonial appearance on nutrient agar, Gram reaction, citrate utilization, indole production, methyl red reaction, Voges–Proskauer reaction and lactose fermentation. The isolates were predominantly Gram-negative rods, with most producing mucoid colonies. The major genera identified were Proteus, Enterobacter, Pseudomonas, Escherichia, Klebsiella, Bordetella, Morganella, Providencia and Citrobacter.All 48 isolates demonstrated the characteristic Gram reaction of Gram-negative bacteria and appeared as Gram-negative rods on microscopic examination. Most isolates produced mucoid colonies on nutrient agar. The principal exception was Pseudomonas spp., for which green pigmentation. The predominance of mucoid colonies was particularly evident among the Enterobacterales, including;Klebsiella, Enterobacter, Escherichia, Proteus, Providencia and Morganella.The colonial and microscopic characteristics provided an initial basis for grouping the isolates before biochemical differentiation. The observation of green colonies among Pseudomonas isolates also provided a useful presumptive characteristic for separating these organisms from the predominantly mucoid Enterobacterales.The biochemical reactions demonstrated considerable phenotypic differentiation among the isolates. Proteus spp. were citrate positive, indole negative and lactose negative, with variable methyl red reactions. Providencia spp. were citrate and indole positive and methyl-red positive but VP and lactose negative. Morganella spp. were citrate negative, indole positive and methyl-red positive, with negative VP and lactose reactions.The Enterobacter and Klebsiella isolates showed a broadly similar biochemical profile, characterized by citrate positivity, indole negativity, VP positivity and lactose fermentation. Escherichia spp. were predominantly indole positive and methyl-red positive but citrate and VP negative. Pseudomonas spp. showed citrate-positive reactions with negative indole, MR, VP and lactose reactions. Citrobacter spp. were citrate positive, indole negative, MR negative, VP negative and lactose positive. These reaction patterns were used collectively for presumptive phenotypic identification rather than relying on a single biochemical test.Across all three compartments, Proteus spp. were the most frequently recovered organisms, accounting for 13 of 48 isolates (27.08%). Enterobacter spp. were the second most abundant (7; 14.58%), followed by Pseudomonas spp. (6; 12.50%). Klebsiella, Escherichia and Bordetella each accounted for 5 isolates (10.42%). Morganella and Providencia each contributed 3 isolates (6.25%), while Citrobacter spp. was the least abundant (1; 2.08%) (Table 3)

3.3. Distribution of Gram-Negative Bacteria

The distributionof Gram-negative bacteria in all three compartments is presented in Table 3.The distribution indicates that several bacterial genera occurred across more than one compartment. Proteus, Enterobacter, Pseudomonas, Klebsiella, Escherichia, Morganella and Providencia were recovered from all the three compartments. Bordetella spp. was recovered exclusively from pig-associated samples, whereas Citrobacter spp. occurred both in the environmental and human compartments. A Pearson chi-square test of independence was used to determine whether the distribution of bacterial genera was significantly associated with the One Health compartment. There was no statistically significant association between bacterial genus and compartment (χ² = 11.25, df = 16, p = 0.79). Thus, although the numerical abundance of individual genera varied between pig-associated, environmental and human-associated samples, the observed distribution did not provide statistical evidence that bacterial genus was dependent on compartment. A separate goodness-of-fit analysis indicated that the overall abundance of the nine bacterial genera was not evenly distributed (χ² = 4.00, df = 2, p = 0.135). This indicates significant variation in the overall abundance of the identified genera, with Proteus spp. occurring considerably more frequently than Citrobacter, Morganella and Providencia spp. The predominance of Proteus across the One-Health compartments suggests possible ecological connectivity, emphasizing the need for integrated surveillance and improved hygiene, waste management, and prudent antimicrobial use within pig production systems.

3.4. Whole-Genome Sequencing and Quality Assessment of Gram-Negative Isolates

The sequencing and assembly metrics for the three One Health compartments are summarized in Table 4 and Table 5. All Gram-negative isolates recovered from the different One Health compartments were subjected to culture-enriched whole-genome shotgun sequencing. The sequenced isolates consisted of 22 from pigs, 14 from the farm environment, and 12 from human-associated samples. The sequencing libraries yielded mean numbers of 17,626,498, 20,085,052, and 23,028,031 high-quality reads for environmental, human-and pig-associated isolates, respectively, with mean Phred quality scores of at least Q30. Following quality-control filtering, 95.50% of the reads from environmental isolates, 96.50% from pig-associated isolates, and 96.00% from human-associated isolates were retained for subsequent bioinformatic analyses. The resulting genome assemblies showed mean N50 values of 21,521 bp for environmental isolates, 7,390 bp for pig-associated isolates, and 4,562 bp for human-associated isolates. The corresponding mean GC contents were 46.30%, 46.04%, and 42.58%, respectively. Overall, the whole-genome sequencing data generated a higher level of genomic resolution and discriminatory power than could be achieved through phenotypic characterization alone, thereby providing more detailed information for differentiating and characterizing the Gram-negative isolates across the One Health compartments.

3.5. Genomic Identification of Sequenced Isolates Recovered From the Pigs

The hierarchical taxonomic analysis of sequenced data obtained from pigs (Figure 1) revealed a diverse bacterial community dominated by Proteobacteria, particularly the class Gammaproteobacteria. Within this group, Enterobacterales was prominent, with Escherichia, Klebsiella, and Enterobacter representing important genera. Escherichia coli and Klebsiella pneumoniae each accounted for approximately 8% of classified sequences, while Enterobacter hormaechei, E. cloacae, and unclassified Enterobacter contributed smaller proportions. Pseudomonas aeruginosa represented approximately 6%, alongside other Pseudomonas sequences. Members of Morganellaceae, including Morganella, Providencia, and Proteus, were also detected, with several Proteus species showing relatively comparable abundances. A distinct Bordetella branch was identified, with Bordetella trematum being the most abundant explicitly labelled taxon at approximately 16%. Overall, the bacterial community demonstrated considerable taxonomic heterogeneity, with explicitly labelled taxa ranging from 2% to 16%. The relatively high abundance of B. trematum contributed to a positively skewed distribution, while numerous taxa occurred at lower proportions of 2–4%. The findings demonstrate that the bacterial population associated with the pigs was not restricted to a single taxonomic group but comprised diverse Gram-negative taxa with varying relative abundances.

3.6. Genomic Identification of Sequenced Isolates Recovered From the Piggery Environment

The taxonomic composition of the bacteria recovered from the piggery environment was assessed using whole-genome shotgun sequencing data and visualized with a Krona hierarchical classification plot (Figure 2). Whole-genome shotgun sequencing of bacteria recovered from the piggery environment revealed a diverse and heterogeneous bacterial community, with sequences predominantly assigned to the phylum Proteobacteria, class Gammaproteobacteria, and order Enterobacterales. The major families identified were Enterobacteriaceae, Morganellaceae, Pseudomonadaceae, and Halomonadaceae. Within Enterobacteriaceae, Escherichia coli, Citrobacter freundii, and Klebsiella variicola were the most prominent taxa, each accounting for approximately 9% of the classified sequences. Other important taxa included Klebsiella pneumoniae (5%), Enterobacter hormaechei (3%), and Enterobacter cloacae (2%). Morganellaceae was represented by Proteus, Providencia, and Morganella, with unclassified Proteus (6%), Proteus rettgeri (5%), P. penneri (5%), P. mirabilis (4%), and Morganella morganii (6%) contributing substantially to the profile. The Pseudomonadaceae family included Pseudomonas qingdaonensis (3%), unclassified Pseudomonas (3%), and Pseudomonas putida (2%). A small proportion of sequences was assigned to Halomonadaceae, represented by Halomonas (1%). Overall, the explicitly identified taxa displayed relative abundances ranging from 1% to 9%, with the three leading taxa collectively accounting for approximately 27% of the profile. The absence of a single overwhelmingly dominant species, together with the presence of numerous low-abundance and unclassified taxa, indicates substantial taxonomic diversity within the piggery environment.

3.7. Genomic Identification of Sequenced Isolates Recovered From the Farm Workers and Humans Living Within the Piggery Environment

The hierarchical taxonomic classification of the sequenced bacterial population is presented as a Krona plot (Figure 3). Culture-enriched whole-genome shotgun sequencing of the metagenomes recovered from farm workers and humans associated with the piggery environment revealed a diverse bacterial population dominated by Proteobacteria, particularly Gammaproteobacteria and the order Enterobacterales. The major families were Enterobacteriaceae and Morganellaceae, with Pseudomonas also represented. Escherichia coli was the most abundant explicitly identified taxon, accounting for approximately 14% of classified sequences, followed by Enterobacter cloacae (9%) and Citrobacter freundii (8%). Other notable taxa included Morganella morganii and unclassified Providencia, each contributing approximately 8%, while unclassified Proteus accounted for 7%. Several Proteus species were identified, including P. penneri and P. faecis (6% each), as well as P. vulgaris and P. mirabilis (2% each). Klebsiella pneumoniae and Pseudomonas taxa occurred at lower proportions. Overall, the explicitly labelled taxa showed relative abundances ranging from 1% to 14%, indicating considerable variation in their representation. The bacterial community was characterized by a predominance of Enterobacterales-associated organisms, particularly members of Enterobacteriaceae and Morganellaceae. However, the presence of numerous additional low-abundance and unclassified taxa demonstrated substantial taxonomic diversity. The distribution indicated that the human-associated bacterial population was not dominated exclusively by a single genus but consisted of multiple bacterial taxa occurring at different relative abundances, highlighting the diverse microbial composition associated with humans within the piggery environment.

3.8. Ncbi Accession Characteristics of Gram-Negative Bacterial Species Isolated From the Pigs

Whole-genome sequencing was used to identify the Gram-negative bacterial isolates recovered from pigs. As presented in Table 6, a total of 12 Gram-negative bacterial isolates were identified to species level. The isolates comprised members of the families Morganellaceae, Enterobacteriaceae, Pseudomonadaceae and Alcaligenaceae, demonstrating considerable taxonomic diversity among the Gram-negative bacteria recovered from the pig samples.The Morganellaceae constituted the largest group, comprising six isolates: Proteus faecis (AOC-BE22), Proteus penneri (AOC-BE23), Proteus mirabilis (AOC-BE24), Proteus vulgaris (AOC-BE25), Providencia rettgeri (AOC-BE26), and Morganella morganii (AOC-BE27). The Enterobacteriaceae were represented by four species, namely Enterobacter hormaechei (AOC-BE28), Enterobacter cloacae (AOC-BE29), Escherichia coli (AOC-BE30), and Klebsiella pneumoniae (AOC-BE31). In addition, Pseudomonas aeruginosa (AOC-BE32) and Bordetella trematum (AOC-BE33) were identified, representing Pseudomonadaceae and Alcaligenaceae, respectively.Among the identified organisms, Proteus spp. were the most represented genus, with four species (P. faecis, P. penneri, P. mirabilis and P. vulgaris), accounting for one-third of the bacterial species identified. Enterobacter was represented by two species, while each of Providencia, Morganella, Escherichia, Klebsiella, Pseudomonas and Bordetella was represented by one species. The whole-genome sequencing data were deposited in the National Center for Biotechnology Information (NCBI) database. Individual isolates were associated with WGS accession-number ranges corresponding to their assembled contigs. The accession series ranged from JCAWQV010000001–JCAWQV010000013 for P. faecis to JCAWQW010000001–JCAWQW010000018 for P. penneri. The number of accessioned contigs varied among isolates, with P. penneri having the highest number of contigs (18), followed by P. faecis, E. hormaechei and P. aeruginosa, each with 13 contigs. P. mirabilis had the lowest number of contigs (6).The variation in the number of contigs among the isolates reflects differences in genome assembly characteristics rather than differences in the number of bacterial species present.

3.9. Ncbi Accession Characteristics of Gram-Negative Bacterial Species Isolated From the Piggery Environment

The identified bacterial species, as shown in Table 7, comprised Klebsiella variicola (AOC-BE1), Klebsiella pneumoniae (AOC-BE2), Citrobacter freundii (AOC-BE3), Escherichia coli (AOC-BE4), Enterobacter hormaechei (AOC-BE5), Enterobacter cloacae (AOC-BE6), Proteus faecis (AOC-BE7), Proteus penneri (AOC-BE8), Proteus mirabilis (AOC-BE9), Proteus vulgaris (AOC-BE10), Providencia rettgeri (AOC-BE11), Morganella morganii (AOC-BE12), Pseudomonas qingdaonensis (AOC-BE13), and Pseudomonas putida (AOC-BE14).The Enterobacteriaceae group comprised six species, namely K. variicola, K. pneumoniae, C. freundii, E. coli, E. hormaechei, and E. cloacae. The Morganellaceae group comprised six species, including P. faecis, P. penneri, P. mirabilis, P. vulgaris, P. rettgeri, and M. morganii. The remaining two isolates were P. qingdaonensis and P. putida, both belonging to the genus Pseudomonas. Thus, the isolates demonstrated representation from three major bacterial lineages within the Gram-negative bacterial community recovered from the piggery environment.

3.10. NCBI Accession Characteristics of the Human-Associated Bacterial Isolates

The bacterial species in human-associated bacterial isolates identified through whole-genome sequencing, as shown in Table 8, comprised Enterobacter cloacae (AOC-BE43), Enterobacter roggenkampii (AOC-BE44), Escherichia coli (AOC-BE45), Citrobacter freundii (AOC-BE46), Klebsiella pneumoniae (AOC-BE47), Proteus faecis (AOC-BE48), Proteus penneri (AOC-BE49), Proteus vulgaris (AOC-BE50), Proteus mirabilis (AOC-BE51), Providencia rettgeri (AOC-BE52), Morganella morganii (AOC-BE53), and Pseudomonas qingdaonensis (AOC-BE54). The isolates represented 12 bacterial species distributed across several genera of Gram-negative bacteria recovered from humans within the piggery production system. The family Enterobacteriaceae comprised five species, namely E. cloacae, E. roggenkampii, E. coli, C. freundii, and K. pneumoniae. The Morganellaceaegroup comprised six species, including P. faecis, P. penneri, P. vulgaris, P. mirabilis, P. rettgeri, and M. morganii. The remaining isolate, P. qingdaonensis (AOC-BE54), belonged to the genus Pseudomonas. Thus, the sequenced isolates demonstrated considerable taxonomic diversity, with representatives of three major Gram-negative bacterial lineages recovered from the pig production system. The whole-genome assemblies were deposited in the National Center for Biotechnology Information (NCBI) database under the corresponding contig accession ranges.

3.11. Distribution of Bacterial Species Across the One Health Compartments

The distribution of bacterial species across the three compartments is presented in Table 9, and demonstrated considerable taxonomic overlap. A total of 17 bacterial species were identified, with the environmental compartment showing the highest species richness (14 species; 82.4%), compared with 12 species (70.6%) each in the pig-associated and human-associated compartments. Nine species (52.9%) were common to all three compartments, namely;Proteus faecis, P. penneri, P. mirabilis, P. vulgaris, Providencia rettgeri, Morganella morganii, Enterobacter cloacae, Escherichia coli, and Klebsiella pneumoniae. This widespread occurrence indicates substantial sharing of bacterial taxa among pigs, the environment, and humans. Several species showed compartment-specific distributions. Pseudomonas aeruginosa and Bordetella trematum were restricted to the pig-associated compartment, whereas Enterobacter roggenkampii occurred exclusively in the human-associated compartment. Klebsiella variicola and Pseudomonas putida were restricted to the environmental compartment, while Citrobacter freundii and Pseudomonas qingdaonensis occurred in both environmental and human-associated samples. Statistical analysis using Cochran's Q test showed no significant difference in the proportion of bacterial species represented across the three compartments (Q = 1.14, df = 2, p = 0.57). Although the environmental compartment recorded the greatest numerical species richness, this difference was not statistically significant at the 5% level. Pairwise Jaccard similarity analysis further demonstrated substantial taxonomic overlap. The pig-associated and environmental compartments shared 10 species, producing a Jaccard coefficient of 0.66, while pig-associated and human-associated compartments shared nine species (J = 0.60). The highest similarity occurred between environmental and human-associated compartments, which shared 11 species (J = 0.73). Overall, the findings indicate broad distribution and substantial taxonomic connectivity of bacteria across the pig production system, with the environment potentially serving as an important interface linking animal and human-associated bacterial populations.